Does liver resection/transplantation affect respiratory induced liver motion in patients with hepatocellular carcinoma?

Abstract The purpose of this study was to evaluate the changes in magnitude of three‐dimensional (3D) liver motion after liver resection/transplantation in patients with hepatocellular carcinoma (HCC) using four‐dimensional (4D)‐computed tomography (CT) images. From January 2012 to April 2016, 74 HCC patients underwent 4D‐CT scans under a free‐breathing state to assess respiratory liver motion. Of the 74 patients, 40 did not have a liver resection/transplantation (Group A), 34 with liver resection/transplantation. 15 underwent major or minor resection in the right liver lobe (Group B), 14 underwent major or minor resection in the left liver lobe (Group C), and five underwent liver transplantation (Group D). The 4D‐CT images were sorted into 10 image series according to the respiratory phase from the end inspiration to the end expiration, and then transferred to treatment planning software. All liver contours were drawn by a single physician and confirmed by a second. Liver relative coordinates were automatically generated to calculate liver respiratory motion in different axial directions and compiled into a single composite image. Differences in respiratory liver motion were assessed using one‐way ANOVA. The average liver respiratory motion in the cranial‐caudal direction and 3D magnitude were 10.46 ± 2.78 mm (range, 5.60–18.80 mm) and 11.74 ± 2.65 mm (range, 7.45–20.79 mm) for patients without liver resection/transplantation, and 7.74 ± 2.79 mm (range, 2.20–12.90 mm) and 9.07 ± 2.38 mm (range, 4.79–14.08 mm) for posthepatectomy/post‐transplant patients respectively. There were significant differences between Group A and B, Group A and C, Group A and D. However, there were no significant differences among Group B, C, and D. Liver resection/transplantation greatly affected respiratory‐induced liver motion in patients with HCC. We, therefore, recommend discriminatory internal target volume (ITV) determination for patients with or without liver resection/transplantation undergoing external radiotherapy for hepatic tumors while respiratory motion management is unavailable.


| INTRODUCTION
Hepatocellular carcinoma (HCC) is a highly prevalent and lethal neoplasia, 1 comprising the majority of primary liver cancers worldwide (70-90%). An estimated 782,500 new liver cancer cases and 745,500 deaths occurred worldwide in 2012 due to HCC, with China alone accounting for approximately 50% of the total number of cases and deaths. 2 The preferred treatments for HCC are surgical resection and percutaneous destruction methods (uni-and multipolar radiofrequency, microwave, cryotherapy, and electroporation). In selected patients, liver transplantation is the best treatment option for small HCC with severe liver cirrhosis. 3 Curative therapies (resection, transplantation, and ablation) can improve survival in patients diagnosed at an early stage of HCC and offer a potential long-term cure. 1,4 However, metastasis is the major risk factor of HCC, which impacts longterm survival of patients with posthepatectomy HCC, and contributes to the high recurrence rate. 5,6 Post-transplant HCC recurrence is reported in up to 25% of cases and drastically affects patient survival. [7][8][9][10] External beam radiotherapy (EBRT) is widely used for HCC in Asia, 11 and when used in combination with hepatic arterial embolization, is a promising treatment. 12 In addition, with current advancements in precision radiotherapy, stereotactic body radiation therapy (SBRT) has also become a promising alternative treatment for patients with primary or recurrent small HCC who are considered unsuitable for surgical resection or local ablative therapy. 13,14 EBRT may play an important role in preventing post-transplant or postoperative recurrence of and/or metastasis from HCC. 13,[15][16][17][18] Patients with unresectable but limited HCC recurrence may undergo EBRT, but the hepatic tumors move during EBRT due to respiratory-induced liver motion. In order to avoid both inadequate tumor coverage and unnecessary liver parenchyma irradiation, it is crucial to determine the internal target volume (ITV). The ITV boundary range primarily relies upon respiration-induced liver motion, and if not properly accounted for, motion of this magnitude could lead to altered dosimetry due to use of a static plan and irradiation of an uncertain volume of normal tissue. 19,20 Inaccurate definitions of the volume of a hepatic tumor and normal tissue could lead to a greater risk of toxicity. Although there are benefits to defining individual ITV, the data are obtained using four-dimensional computed tomography (4D-CT), but the process of contouring each phase is timeconsuming and labor-intensive. The gross target volume (GTV) must be manually contoured to form ITV in all respiratory phases of a 4D scan image. In addition, the 4D-CT technique is not universally available in all radiation oncology centers, and some radiation oncologists may determine the margin ITV based upon their individual experience. In theory, ligament damage and tissue adhesions surrounding a remnant liver may cause a decrease of amplitude in respiratoryinduced liver motion. To date, the impact on ITV margins after liver resection in HCC patients has not been reported. Therefore, in this study, we investigated the differences in liver motion between posttransplant or postoperative recurrence HCC patients and unresectable HCC patients in a free-breathing state to provide a valuable reference for radiation oncologists when determining ITV. with HCC were divided into four groups (described in more detail below) and underwent 4D-CT scans to assess respiratory liver motion.

2.C | 4D-CT image acquisition
4D-CT scans were obtained using a Big Bore CT Scanner (Siemens Somatom CT, Sensation Open; Siemens Healthcare, Munchen Germany). Patients were placed in a supine position with arms raised above the forehead, and were immobilized using a vacuum cushion.
The X-ray tube settings were: 120 KV; 400 mAs; Pitch 0.1; Gantry rotation cycle time 0.5 s; 3 mm reconstructed thickness. The respiratory phase on the respiratory wave was manually adjusted and confirmed by the CT-simulation technician prior to CT image reconstruction. 4D-CT images from respiratory raw data were sorted into a 10 CT image series (CT0~CT90) according to the respiratory cycle, with CT0 being defined as the end inspiration phase and CT50 as the end expiration phase. 21 Datasets for 4D-CT scans were then transferred to Nucletron Oncentra's treatment planning software

2.D | Liver displacement acquisition and analysis
Liver contours were delineated at all CT image phases and then copied manually to a single plan. Nine liver contours of CT10~CT90 were copied onto the CT0 image, and were designated CopyCon-tour10~CopyContour90.
There were 10 liver contours (CopyContour10~CopyContour90 and liver contours of CT0) on the CT0 image. An AP digitally reconstructed radiography image was created to order visualize each phase contour (Fig. 1). The relative coordinates of the liver were automatically generated to calculate the respiratory liver motion in different axial directions.
The position for each liver was expressed using the left-right (LR), cranial-caudal (CC), and anterior-posterior (AP) coordinates of the center of mass (COM) for each 4D-CT bin. Then, the range of respiratory liver motion from the COM of each coordinate was obtained. The maximum range of motion in each axial direction was obtained by subtracting the minimum relative coordinate value from the maximum relative coordinate value. The 3D motion magnitude of the COM was calculated according to the following formula: Variables were expressed as the mean AE standard deviation.

2.E | Statistical analyses
A Chi-square (v 2 ) test was used to compare patient demographics and clinical characteristics between the four patient groups (A-D). The variation between the four groups in the LR, CC, AP, and 3D directions were assessed using a one-way ANOVA test, using Student's ttest to compare breath amplitude of patients with and without liver resection/transplantation, and liver motion by different postoperative time nodes (Table 5). Post Hoc Test was used to perform multiple comparisons of liver motions among the four groups ( Table 4). The T A B L E 1 Patient demographics and clinical characteristics.

3.B | Distribution of CC displacement in the four patient groups
As shown in Fig. 2 The mean difference is significant at the 0.05 level. "Sig."stands for "P value".
T A B L E 5 Comparison and analysis of the respiratory liver motion (mm) in 3D magnitude in patients with liver resection/transplantation at different postoperative periods.

| DISCUSSION
The human body achieves gas exchange with its surrounding environment primarily with respiratory motion, using diaphragmatic muscles for breathing. The diaphragm pulls on the liver via a ligament, which induces liver motion, while some ligaments surrounding the liver that are not attached to the diaphragm limit respiratory-induced liver motion. Thus, ligament function is a critical factor for respiratory-induced liver motion amplitude. Liver resection and liver transplantation can cause detachment of the ligaments involved in liver respiratory motion. A right-sided hepatectomy may cause detachment of, among others, the hepatorenal ligament, the round ligament of the liver, the hepatic falciform ligament, the right coronary ligament, and the right triangle ligament. A left-sided hepatectomy may cause detachment of the round ligament of the liver, the hepatic falciform ligament, the left coronary ligament, the left triangle ligament, and the hepatogastric ligament, among others. During a liver transplant, all peri-hepatic ligaments will be cut. 22 Correspondingly, this study found that respiratory-induced liver motion was smaller in HCC patients with liver resection/transplantation compared to those without liver resection/transplantation. Some researchers believe that using the COM of liver for analysis overly condenses the data and may not be representative of liver motion. In addition, there is also concern that the reproducibility of manually drawing liver contours impacts accuracy. In this study, we explored respiratory-induced liver motion primarily from a macroperspective. In fact, we did attempt the method of "border locations". However, quantitative analysis is difficult for two reasons: (a) The drawing error would become bigger using the "border location" method than the error (<0.2 mm) in the COM method, which leads to weak quantitative accuracy; and (b) The inconsistency of liverinduced "border location" motion may occur. "Border location" motion may not necessarily equal liver motion. 23 The maximum "border location" motion was very difficult to find, but we still considered "border location" an effective and intuitive method, as illustrated in the representative liver motion images in Fig. 1 from a qualitative perspective. A certain drawing error could inevitably exist in this study, and we explored this issue before we initiated the study. The liver contours were drawn five times in the same patient's 4D-CT image at different times by a single radiation oncologist (HY), and then the drawing error was compared. The differences of each coordinate value among the COM of five liver contours drawn in the same patient's 4D-CT image were all less than 0.2 mm, which was deemed acceptable in this study. Therefore, we determined that the liver contours drawn by HY were reproducible, and the drawing error would not impact the accuracy by which COM (and thus motion magnitudes) were determined.
In theory, reduced liver motion can lead to reduced ITV. 24,25 Therefore, it is important to manage and/or account for respiratory liver motion through means such as abdominal compression (AC), 26 which uses a constant force applied to the abdomen to reduce liver motion, respiratory gating techniques 27  All patients in this study had undergone helical tomotherapy in a freebreathing state in our institution. However, due to the lack of 4D-CT equipment in some radiotherapy institutions, radiation oncologists must rely on their own experience to determine the ITV, which is critical for EBRT success in HCC patients with an intrahepatic tumor. Radiation oncologists should consider respiratory-induced liver motion differently for HCC patients with or without liver resection/transplantation, which is crucial in estimating ITV.
In fact, intrahepatic tumor motion is not equal to respiratoryinduced liver motion. 23 Technologies explored by radiation oncolo-  36 They also used 4DCT to quantify multiorgan respirationinduced motion in the abdomen, and found that the average liver motion and liver tumor motion were 7.8 AE 2.6 (range, 3-13) mm and 9.7 AE 5.0 (range, 3-18) mm in CC direction. 36  demonstrated that cine-MRI detected differences in hepatic tumor motion when compared with 4DCT, cine-MRI motion was larger than 4DCT for the CC direction in 50% of patients by a median of 3.0 mm (range, 1.5-7 mm), the AP direction in 44% of patients by a median of 2.5 mm (range, 1-5.5 mm), and LR in 63% of patients by a median of 1.1 mm (range, 0.2-4.5 mm). They considered that the cine-MRI had better time resolution and was better able to capture the extreme positions of the tumor motion than 4DCT as the reason. 37 More studies are required to investigate the phenomenon and identify which is the real liver tumor motion based on different imaging modalities.
4D-CT is helpful to determine the internal target volume, but if 4D-CT is not available then the data of the result in this study could be used along with published margins recipes to determine population-based target volumes. If possible, the motion of the actual tumor, rather than the liver center of mass motion, should be used for target volume generation.

| CONCLUSIONS
Liver resection/transplantation greatly affects respiratory-induced liver motion in patients with HCC. We determined that the respiratory-induced liver motion in HCC patients with liver resection/transplantation was smaller than that in HCC patients without liver resection/transplantation. Therefore, we recommend discriminatory ITV determination in patients with or without liver resection/ transplantation undergoing external radiotherapy for hepatic tumors while respiratory motion management is unavailable.

CONFLI CT OF INTEREST
The authors declare no conflicts of interest.